Skip to main content
GenosOS runs on your hardware. Your conversations are encrypted at rest with AES-256-GCM and never leave your machine. There are no config files to edit and no dashboard to learn — you configure everything by talking to the assistant.

​
Why GenosOS?

Most AI assistants share three problems: they are stateless (they forget you between sessions), cloud-only (your conversations live on someone else’s server), and technically demanding (environment variables, config files, deployment scripts). GenosOS is built to solve all three.
Typical AI assistantGenosOS
MemoryStateless — resets each sessionPersistent memory with semantic search
Data locationCloud-hostedYour hardware only
ConfigurationConfig files, dashboardsNatural language conversation
EncryptionOptional or noneAES-256-GCM, mandatory

​
Core differentiators

Runs on your hardware. The gateway binds to loopback by default. Your data stays on your machine. Encrypted by default. AES-256-GCM encryption at rest (PBKDF2 100K iterations). Every config write is encrypted automatically — the agent does not need to know it is happening. Remembers you. Structured memory with semantic search. Your assistant knows your preferences, your projects, and your context across sessions. TOON compaction reduces token usage by ~40% while preserving meaning. Works across channels. One assistant, connected to all your messaging apps. Messages from any channel land in the same conversation with full memory and context. Configured by talking. The agent understands 30 configuration actions, validated by 190 server-side blueprints. Say what you want — the agent handles the how.

​
Configured by talking

No forms. No JSON editing. Describe what you want and the agent makes it happen. 
You:  "Connect my WhatsApp"
      → QR code appears, you scan it, done.

You:  "Only let my contacts message me"
      → Security policies configured automatically.

You:  "Create an assistant for my dental clinic"
      → Specialized agent deployed with appointment scheduling
        and patient communication — in one conversation.

You:  "Harden security"
      → Fortress Mode enabled: Keychain storage, buffer zeroing,
        SQLite hardening, vault auto-lock, rate limiting.

​
Supported channels

Core messaging:
ChannelLibraryStatus
WhatsAppBaileysProduction
TelegramgrammYProduction
SlackBoltProduction
Discorddiscord.jsProduction
Signalsignal-utilsProduction
iMessagemacOS nativeProduction
Voice:
ChannelTechnologyStatus
Voice callsTwilio/Telnyx + OpenAI RealtimeProduction
Realtime callOpenAI Realtime (bidirectional)Experimental
Local TTSKokoro (on-device, no cloud)Production
Additional: Google Chat, Microsoft Teams, Matrix, Nostr, Twitch, LINE, WebChat, and more.

​
AI providers

Three curated, fully-tested providers with tool calling support:
ProviderDefault modelBoost model
AnthropicClaude Sonnet 4.6Claude Opus 4.6
OpenAIGPT-5.4o3
GeminiGemini 2.5 ProGemini 3 Pro
Pick a provider, paste your API key, and you’re done. The assistant auto-escalates to the boost model when you express urgency or need for deeper reasoning — no manual switching required.

​
Runtime

GenosOS is built on Bun (>=1.2.0) and written in pure JavaScript (ES2024+). There is no TypeScript compilation and no build step for development. Bun provides native SQLite, fast startup, and instant ESM resolution.

​
Explore the docs

Quick start

Get GenosOS running in under five minutes — from install to first conversation.

Connect channels

Link WhatsApp, Telegram, Slack, Discord, Signal, and more.

AI providers

Choose and configure Anthropic, OpenAI, or Gemini.

Security and vault

Understand encryption, Fortress Mode, and WebAuthn/Touch ID.